With over 5,000 dams submitted to the Brazilian Dam Safety Plan, and two recent collapse incidents causing more than 300 deaths and major environmental damage, Brazilian authorities have tightened inspection and upkeep requirements in the country.
“Many accident reports show that problems were not detected by instrumentation but by visual observation. Drones can help, but due to the large dimensions of these structures we need much longer flight times,” explains Carlos Eduardo Cavaco Gomes, director from EPH Engenharia.
Some of the dams are so large that they would require months of battery-powered drone flights to fully scan their surfaces. The new HYCOPTER fuel cell drones can scan 6 times more surface area, making large-scale dam inspections faster to complete.
Designed in Austin, Texas- the HYCOPTER hydrogen fuel cell drone is also equipped with a new data acquisition software successfully tested last month at a COPEL hydropower dam site in Brazil.
Data collected by HYCOPTER links to H3 Zoom.AI, a new AI-assisted inspection platform designed to speed up expert validation processes. Many defect types can now be reported by H3 Dynamics’ machine learning engine. “We can now improve the coverage area, agility, scalability, frequency and quality of visual dam inspections,” said Mr Gomes.
H3 Dynamics has already deployed its AI solution on over 150 tall and complex high-rise buildings in Singapore. Given similarities in materials and structures, the accumulated knowledge can now also be applied to dams.
Furthermore, H3 Dynamics is also planning to install a network of its emergency DBX drone stations at high risk facilities to provide rescue assistance in case of disasters. DBX drone stations were designed for deployment in remote, off-grid environments – where many of these installations are located.
The 91,000 dams in the United States are also a costly safety concern. The cost of rehabilitating life-threatening dams is estimated at $45 billion, and the cost of fixing those in need of repair reaches nearly $71 billion.